Eliminating Primacy Bias in Online Reinforcement Learning by Self-Distillation.

Excessive invalid explorations at the beginning of training lead deep reinforcement learning process to fall into the risk of overfitting, further resulting in spurious decisions, which obstruct agents in the following states and explorations. This phenomenon is termed primacy bias in online reinforcement learning. This work systematically investigates the primacy bias in online reinforcement learning, discussing the reason for primacy bias, while the characteristic of primacy bias is also analyzed. Besides, to learn a policy generalized to the following states and explorations, we develop an online reinforcement learning framework, termed self-distillation reinforcement learning (SDRL), based on knowledge distillation, allowing the agent to transfer the learned knowledge into a randomly initialized policy at regular intervals, and the new policy network is used to replace the original one in the following training. The core idea for this work is distilling knowledge from the trained policy to another policy can filter biases out, generating a more generalized policy in the learning process. Moreover, to avoid the overfitting of the new policy due to excessive distillations, we add an additional loss in the knowledge distillation process, using L2 regularization to improve the generalization, and the self-imitation mechanism is introduced to accelerate the learning on the current experiences. The results of several experiments in DMC and Atari 100k suggest the proposal has the ability to eliminate primacy bias for reinforcement learning methods, and the policy after knowledge distillation can urge agents to get higher scores more quickly.

Serotonergic neurotransmission mediated cognitive dysfunction in two mouse models of sepsis-associated encephalopathy.

BACKGROUND: Patients with sepsis-associated encephalopathy (SAE) often exhibit cognitive impairments. Despite this, the underlying mechanisms of SAE remain largely unexplored. Here, we explored the role of serotonergic neurotransmission in cognitive dysfunction of two mouse models of SAE. METHODS: The mouse models of SAE were established by injection of lipopolysaccharide (LPS, 10 mg/kg, intraperitoneal) and cecal ligation puncture (CLP) respectively. Barnes maze, new object recognition test and open field test were used to evaluate the effects of fluoxetine (selective serotonin reuptake inhibitor) and cyproheptadine (nonselective 5-HT2 receptor antagonist) on cognition and motor activity of mice. Additionally, WAY100635 (5-HT1A receptor antagonist) was co-administered with fluoxetine to explore the mechanism underlying effect of fluoxetine on cognitive impairments of SAE. Enzyme-linked immunosorbent assay (ELISA) was performed to determine 5-HT levels in hippocampus, brainstem and frontal lobe of experimental groups. RESULTS: Both LPS-induced sepsis and CLP induced sepsis resulted in a notable learning deficit. Fluoxetine ameliorated, while cyproheptadine aggravated, cognitive impairment in two classic mouse models of SAE. The cognition-enhancing effect of fluoxetine is reversed by WAY100635. Decreased 5-HT levels in hippocampus, brainstem and frontal lobe were observed in LPS septic model and CLP septic model. Notably, both fluoxetine and cyproheptadine significantly increased 5-HT levels in those brain regions in LPS septic model. Additionally, fluoxetine significantly increased 5-HT levels in frontal lobe of CLP septic model. CONCLUSIONS: Our study demonstrated that serotonergic neurotransmission plays a significant role in mechanisms underlying cognitive impairment in SAE. These findings contribute to identification of novel targets to prevent and arrest cognitive impairment in SAE.

Unbiased organism-agnostic and highly sensitive signal peptide predictor with deep protein language model.

Signal peptides (SPs) are essential to target and transfer transmembrane and secreted proteins to the correct positions. Many existing computational tools for predicting SPs disregard the extreme data imbalance problem and rely on additional group information of proteins. Here we introduce Unbiased Organism-agnostic Signal Peptide Network (USPNet), an SP classification and cleavage-site prediction deep learning method. Extensive experimental results show that USPNet substantially outperforms previous methods on classification performance by 10%. An SP-discovering pipeline with USPNet is designed to explore unprecedented SPs from metagenomic data. It reveals 347 SP candidates, with the lowest sequence identity between our candidates and the closest SP in the training dataset at only 13%. In addition, the template modeling scores between candidates and SPs in the training set are mostly above 0.8. The results showcase that USPNet has learnt the SP structure with raw amino acid sequences and the large protein language model, thereby enabling the discovery of unknown SPs.

Pink1 deficiency enhances neurological deficits and inflammatory responses after intracerebral hemorrhage in mice.

Pink1 (PTEN-induced putative kinase 1) is a protein associated with maintaining mitochondrial function and integrity and has been reported to mediate neurodegeneration and neuroinflammation. While the role of Pink1 in intracerebral hemorrhage (ICH)-related neurological deficits and inflammatory responses is not deciphered. Congenic blood was transfused into the left corpus striatum to construct the ICH model in C57/BL6 wild-type (WT) and Pink1-/- mice. The relative expression of Pink1, monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein (MIP)-2, tumor necrosis factor (TNF)-α, interleukin (IL)-1ß, Cd86, nitric oxide synthase 2 (Nos2), Cd206, arginase 1 (Arg-1), and IL-10 was detected with qRT-PCR, Western blotting, or ELISA. Mouse neurological deficit scores (mNSS) and water content were detected, and an open-field test was performed to assay anxiety-like behavior. Remarkably decreased Pink1 expression and increased MIP-2, IL-1ß, MCP-1, and TNF-α expression were observed after 12 â€‹h, 24 â€‹h, 48 â€‹h, 72 â€‹h, and 7 â€‹d post-ICH induction in the ipsilateral injury hemispheres. Pink1 deficiency could further up-regulate mNSS scores, brain water content, MIP-2, MCP-1, IL-1ß, and TNF-α in the ipsilateral injury hemispheres. On the other hand, Pink1 deficiency could decrease the number of center cross, the velocity, and the total distance traveled in open field test. Pink1 deficiency could further up-regulate the mRNA levels of pro-inflammatory (M1) molecules (Cd86, Nos2), and down-regulate the relative expression of anti-inflammatory (M2) molecules (Cd206, Arg-1, and IL-10). Pink1 deficiency deteriorates neurological deficits and inflammatory responses after ICH, which can be considered as a treatment target.

Design method of dual-band synchronous zoom microscope optical system based on coaxial Kohler illumination.

The intrinsic properties of the observed object are closely related to its spectral information, to extend the imaging spectral range of a continuous zoom microscope to obtain more detailed intrinsic properties of the object, this paper proposes a design method of dual-band simultaneous zoom microscope optical system based on the coaxial Koehler uniform illumination. First, the imaging principle of the dual-band simultaneous zoom microscope optical system is theoretically analyzed, and we propose to split the front fixed group of the zoom system into a collimation lens group and a converging lens group to realize the compact design of the system. Then, two different rear fixed groups are used to correct the residual aberration, and a method for solving the initial structure of the dual-band simultaneous zoom microscope optical system is proposed. Finally, a dual-band synchronous zoom microscope optical system is designed using the method proposed in this paper. The design results show that the imaging magnification of the visible (VIS) band is -0.4 to -4.0, the simultaneous imaging magnification ranges are -0.4 to -0.8 in the VIS and short-wave infrared (SWIR) bands, and the magnification difference of its simultaneous zoom imaging is less than 1.25%. In addition, the system has the advantages of good imaging quality, clever design of coaxial illumination, and compact structure, thus verifying the feasibility of the design method.

Intelligent Recognition Using Ultralight Multifunctional Nano-Layered Carbon Aerogel Sensors with Human-Like Tactile Perception.

Humans can perceive our complex world through multi-sensory fusion. Under limited visual conditions, people can sense a variety of tactile signals to identify objects accurately and rapidly. However, replicating this unique capability in robots remains a significant challenge. Here, we present a new form of ultralight multifunctional tactile nano-layered carbon aerogel sensor that provides pressure, temperature, material recognition and 3D location capabilities, which is combined with multimodal supervised learning algorithms for object recognition. The sensor exhibits human-like pressure (0.04-100 kPa) and temperature (21.5-66.2 °C) detection, millisecond response times (11 ms), a pressure sensitivity of 92.22 kPa-1 and triboelectric durability of over 6000 cycles. The devised algorithm has universality and can accommodate a range of application scenarios. The tactile system can identify common foods in a kitchen scene with 94.63% accuracy and explore the topographic and geomorphic features of a Mars scene with 100% accuracy. This sensing approach empowers robots with versatile tactile perception to advance future society toward heightened sensing, recognition and intelligence.

Synaptotagmin-11 regulates immune functions of microglia in vivo.

Membrane trafficking pathways mediate key microglial activities such as cell migration, cytokine secretion, and phagocytosis. However, the underlying molecular mechanism remains poorly understood. Previously, we found that synaptotagmin-11 (Syt11), a non-Ca2+ -binding Syt associated with Parkinson's disease (PD) and schizophrenia, inhibits cytokine release and phagocytosis in primary microglia. Here we reported the in vivo function of Syt11 in microglial immune responses using an inducible microglia-specific Syt11-conditional-knockout (cKO) mouse strain. Syt11-cKO resulted in activation of microglia and elevated mRNA levels of IL-6, TNF-α, IL-1ß, and iNOS in various brain regions under both resting state and LPS-induced acute inflammation state in adult mice. In a PD mouse model generated by microinjection of preformed α-synuclein fibrils into the striatum, a reduced number of microglia migrated toward the injection sites and an enhanced phagocytosis of α-synuclein fibrils by microglia were found in Syt11-cKO mice. To understand the molecular mechanism of Syt11 function, we identified its direct binding proteins vps10p-tail-interactor-1a (vti1a) and vti1b. The linker domain of Syt11 interacted with both proteins and a peptide derived from it competitively inhibited the interaction of Syt11 with vti1a/vti1b in vitro and in cells. Importantly, application of this peptide induced more cytokine secretion in wild-type microglia upon LPS treatment, phenocopying defects in Syt11 knockdown cells. Altogether, we propose that Syt11 inhibits microglial activation in vivo and regulates cytokine secretion through interactions with vti1a and vti1b.

Design of the dual-band shared-aperture asynchronous zoom optical system using focus tunable lenses.

To meet the increasing demand of the current market and the diversity of application scenarios, combine the zoom system and the multi-band shared-aperture system, and fully harness their respective advantages, this paper proposes a dual-band shared-aperture asynchronous zoom optical system using focus tunable lenses (FTLs). To address the lack of available patents for such systems, we designed a sub-system simultaneous iterative optimization algorithm to calculate the initial structure parameters. This synchronous iterative optimization approach can strengthen the connection between sub-systems and compensate for the shortcomings of current mainstream design methods. The initial structure constructed in this way has a good performance in terms of structural stability and optimization potential. Based on these methods, we successfully designed an optical system that can work in both VIS and NIR bands, and the two sub-systems can zoom independently. The design results possess good performance in terms of distortion control, aberration correction, and volume control.

p55γ degrades RIP3 via MG53 to suppress ischaemia-induced myocardial necroptosis and mediates cardioprotection of preconditioning.

AIMS: Regulated necrosis (necroptosis) and apoptosis are important biological features of myocardial infarction, ischaemia-reperfusion (I/R) injury, and heart failure. However, the molecular mechanisms underlying myocardial necroptosis remain elusive. Ischaemic preconditioning (IPC) is the most powerful intrinsic cardioprotection against myocardial I/R injury. In this study, we aimed to determine whether IPC suppresses I/R-induced necroptosis and the underlying molecular mechanisms. METHODS AND RESULTS: We generated p55γ transgenic and knockout mice and used ligation of left anterior descending coronary artery to produce an in vivo I/R model. The effects of p55γ and its downstream molecules were subsequently identified using mass spectroscopy and co-immunoprecipitation and pulldown assays. We found that p55γ expression was down-regulated in failing human myocardium caused by coronary heart disease as well as in I/R mouse hearts. Cardiac-specific p55γ overexpression ameliorated the I/R-induced necroptosis. In striking contrast, p55γ deficiency (p55γ-/-) and cardiac-specific deletion of p55γ (p55γc-KO) worsened I/R-induced injury. IPC up-regulated p55γ expression in vitro and in vivo. Using reporter and chromatin immunoprecipitation assays, we found that Hif1α transcriptionally regulated p55γ expression and mediated the cardioprotection of IPC. IPC-mediated suppression of necroptosis was attenuated in p55γ-/- and p55γc-KO hearts. Mechanistically, p55γ overexpression decreased the protein levels of RIP3 rather than the mRNA levels, while p55γ deficiency increased the protein abundance of RIP3. IPC attenuated the I/R-induced up-regulation of RIP3, which was abolished in p55γ-deficient mice. Up-regulation of RIP3 attenuated the p55γ- or IPC-induced inhibition of necroptosis in vivo. Importantly, p55γ directly bound and degraded RIP3 in a ubiquitin-dependent manner. We identified MG53 as the E3 ligase that mediated the p55γ-induced degradation of RIP3. In addition, we also found that p55γ activated the RISK pathway during IPC. CONCLUSIONS: Our findings reveal that activation of the MG53-RIP3 signal pathway by p55γ protects the heart against I/R-induced necroptosis and underlies IPC-induced cardioprotection.

Optical system design of double-sided telecentric microscope with high numerical aperture and long working distance.

The working distance of the high numerical aperture visible video microscope is extremely short, which greatly limits its application scenarios. To solve this problem, this paper proposes an unobstructed design method of double-sided telecentric microscope with high numerical aperture and long working distance. First, aiming at the obstruction problem of the image-side telecentric catadioptric microscope objective, the structure of the catadioptric optical system is improved. Then, the aspheric design method based on the best aberration compensation is analyzed theoretically to better correct the primary aberration of the high-numerical aperture microscope objective. Finally, a double-sided telecentric microscope optical system with a numerical aperture (NA) of 0.8 and a working distance of 10.0 mm was designed, which is composed of a spherical reflector, a beam splitter plate, a collimating lens group, and an image-side telecentric eyepiece optical system. The design results show that the imaging resolution of this high numerical aperture video microscope is as high as 0.42 µm, and the microscope has a magnification of about 220× for the image with 1080P (1920 × 1080 pixels) resolution. This double-sided telecentric microscope has the advantages of a large field of view, compact structure, good stray light suppression ability, and manufacturability, and has high practical value in the field of high-precision measurement and detection.

Semi-Supervised Detection Model Based on Adaptive Ensemble Learning for Medical Images.

Introducing deep learning technologies into the medical image processing field requires accuracy guarantee, especially for high-resolution images relayed through endoscopes. Moreover, works relying on supervised learning are powerless in the case of inadequate labeled samples. Therefore, for end-to-end medical image detection with overcritical efficiency and accuracy in endoscope detection, an ensemble-learning-based model with a semi-supervised mechanism is developed in this work. To gain a more accurate result through multiple detection models, we propose a new ensemble mechanism, termed alternative adaptive boosting method (Al-Adaboost), combining the decision-making of two hierarchical models. Specifically, the proposal consists of two modules. One is a local region proposal model with attentive temporal-spatial pathways for bounding box regression and classification, and the other one is a recurrent attention model (RAM) to provide more precise inferences for further classification according to the regression result. The proposal Al-Adaboost will adjust the weights of labeled samples and the two classifiers adaptively, and the nonlabel samples are assigned pseudolabels by our model. We investigate the performance of Al-Adaboost on both the colonoscopy and laryngoscopy data coming from CVC-ClinicDB and the affiliated hospital of Kaohsiung Medical University. The experimental results prove the feasibility and superiority of our model.

Cooper Pair Excitation Mediated by a Molecular Quantum Spin on a Superconducting Proximitized Gold Film.

Breaking a correlated pair in a superconductor requires an even number of fermions providing at least twice the pairing energy Δ. Here, we show that a single tunneling electron can also excite a pair breaking excitation in a proximitized gold film in the presence of magnetic impurities. Combining scanning tunneling spectroscopy with theoretical modeling, we map the excitation spectrum of an Fe-porphyrin molecule on the Au/V(100) proximitized surface into a manifold of entangled Yu-Shiba-Rusinov and spin excitations. Pair excitations emerge in the tunneling spectra as peaks outside the spectral gap only in the strong coupling regime, where the presence of a bound quasiparticle in the ground state ensures the even fermion parity of the excitation. Our results unravel the quantum nature of magnetic impurities on superconductors and demonstrate that pair excitations unequivocally reveal the parity of the ground state.

Optimizing The Timing of Stereotactic Minimally Invasive Drainage for Hypertensive Intracerebral Hemorrhage.

INTRODUCTION: Intracerebral hemorrhage is a high-risk pathological event that is associated with formidable morality rates. Here, our objective was to perform a retrospective study to determine the best timing for drainage using physiological data on patients who received drainage at different timings. METHODS: In this retrospective study, we reviewed 198 patients with hypertensive cerebral hemorrhage who underwent stereotactic drainage at the conventional timing (surgery within 12 h of admission; control group) and 216 patients who underwent stereotactic drainage at a customized surgical timing (elective group). Follow-ups were performed at 3 and 6 months after surgery. RESULTS: The clinical indicators, including prognosis, hematoma clearance, recurrent hemorrhage, intracerebral infection, pulmonary infection, deep venous thrombosis, gastrointestinal hemorrhage, National Institutes of Health Stroke Scale scores, and matrix metallopeptidase 2 and 9 levels, were compared between the control and elective groups. Our data indicated that the elective group had significantly better prognosis compared to the control group (p = 0.021), with a higher rate of hematoma clearance (p = 0.004) and a lower rate of recurrent hemorrhage (p = 0.018). The total occurrence rate of post-surgery complications was also lower for the elective group (p = 0.026). NIHSS scores and serum MMP2/9 levels of the elective group were lower than those of the control group. CONCLUSIONS: Customized timing of stereotactic drainage may be superior to conventional fixed timing (within 12 h post-hemorrhage) in reducing post-surgery complications and promoting recovery, which supports the potential use of customized timing of stereotactic minimally invasive drainage as a new convention in clinics.

Impact of aqueous environments on hydrogen peroxide activation by manganese oxides: Kinetics and the critical role of bicarbonate.

MnO2 activating H2O2 is a promising way in the field of advanced oxidation processes (AOPs) and in situ chemical oxidation (ISCO) to remove contaminants. However, few studies have focused on the influence of various environmental conditions on the performance of MnO2-H2O2 process, which restricts the application in real world. In this study, the effect of essential environmental factors (ionic strength, pH, specific anions and cations, dissolved organic matter (DOM), SiO2) on the decomposition of H2O2 by MnO2 (ε-MnO2 and ß-MnO2) were investigated. The results suggested that H2O2 degradation was negatively correlated with ionic strength and strongly inhibited under low pH conditions and with phosphate existence. DOM had a slight inhibitory effect while Br-, Ca2+, Mn2+ and SiO2 placed negligible impact on this process. Interestingly, HCO3- inhibited the reaction at low concentrations but promoted H2O2 decomposition at high concentrations, possibly due to the formation of peroxymonocarbonate. This study may provide a more comprehensive reference for potential application of H2O2 activation by MnO2 in different water systems.

Double-sided telecentric zoom optical system using adaptive liquid lenses.

In this paper, the adaptive liquid (AL) lenses are introduced into the double-sided telecentric zoom system, which could greatly decrease the mechanical motion group compared with the traditional zoom system, and only one movable aperture stop (STO) is retained. Firstly, this paper derived the Gaussian brackets used in this system, and we found the appropriate screening method to get the suitable initial structure parameters from the solution space. Then we used the lens module design method to create the initial system. Finally, we used CODEV to further optimize the system, and we got an excellent design result, which controlled the telecentricity of both sides within 0.1°, the distortion was controlled within 0.5%, and the MTF of each zoom configuration above 0.3. This optical system has high application potential and value in the field of precision machine vision. The design method proposed in this article can effectively solve the problem that the zoom system using adaptive liquid lenses lacks the initial structure.

Lateral Transfer Learning for Multiagent Reinforcement Learning.

Some researchers have introduced transfer learning mechanisms to multiagent reinforcement learning (MARL). However, the existing works devoted to cross-task transfer for multiagent systems were designed just for homogeneous agents or similar domains. This work proposes an all-purpose cross-transfer method, called multiagent lateral transfer (MALT), assisting MARL with alleviating the training burden. We discuss several challenges in developing an all-purpose multiagent cross-task transfer learning method and provide a feasible way of reusing knowledge for MARL. In the developed method, we take features as the transfer object rather than policies or experiences, inspired by the progressive network. To achieve more efficient transfer, we assign pretrained policy networks for agents based on clustering, while an attention module is introduced to enhance the transfer framework. The proposed method has no strict requirements for the source task and target task. Compared with the existing works, our method can transfer knowledge among heterogeneous agents and also avoid negative transfer in the case of fully different tasks. As far as we know, this article is the first work denoted to all-purpose cross-task transfer for MARL. Several experiments in various scenarios have been conducted to compare the performance of the proposed method with baselines. The results demonstrate that the method is sufficiently flexible for most settings, including cooperative, competitive, homogeneous, and heterogeneous configurations.

Prediction Model of VCA Formed by the Packing of Hybrid Lithological Coarse Aggregates Used in SMA.

The voids in coarse aggregate (VCA) is an important volumetric index in the mineral aggregate gradation design of stone matrix asphalt (SMA) mixtures. To explore the law of variation for VCA formed by the packing of basalt and lime coarse aggregates, a uniform design method and vibrating compaction tests were used to establish the prediction model. Based on the test results and stepwise regression analysis, a reliable prediction model of VCA was obtained. There is a multiple nonlinear relationship between the VCA and the proportion of each coarse aggregate in the mixture. Regardless of the type of coarse aggregates used, the rule of VCA with different forms of aggregate gradation curves has universal significance. This conclusion can help to determine the aggregate gradation in the design of SMA mixtures.

Design of a Concentric Multi-Scale Zoom Optical System Based on Wide Object Distance and High-Precision Imaging.

To effectively balance the trade-off between a large field of view (FOV) and high resolution of an optical system, as well as to solve the problem of image stitching misalignment after focusing, firstly, this paper conducts a theoretical analysis of the design principle of the concentric multi-scale optical system and the causes of image stitching misalignment after focusing. Secondly, the design idea of using a combination structure of a two-layer front concentric imaging group and an image-space telecentric relay imaging array and then a joint full-motion zoom relay imaging system is proposed. Finally, an image-space telecentric two-step zoom concentric multi-scale optical system with a 7 × 7 relay imaging array is designed. The FOV of this optical system is 60° × 45°; the focal lengths are 50 mm and 100 mm for the center channel and 50 mm for the other channels. This concentric multi-scale zoom system has the advantages of both high-precision imaging stitching with a wide object distance and high-resolution imaging, which makes up for the defects of the conventional concentric multi-scale optical system, making it a promising application in the fields of aviation and security.

High-precision calibration to zoom lens of optical measurement machine based on FNN.

We proposed a calibration method for high-precision zoom lenses of optical measurement machines based on Fully Connected Neural Network (FNN), using a 5-layer neural network instead of a camera calibration model, to achieve continuous calibration of zoom lenses at any zoom setting by calibrating typical zooms. From the experimental verification, the average calibration error of this method is 9.83×10-4mm and the average measurement error at any zoom setting is 0.01317mm. The overall calibration precision is better than that of Zhang's calibration method and can meet the application requirements of a high-precision optical measurement machine. The method proposed in this paper provided a new solution and a new idea for the calibration of zoom lenses, which can be widely used in the fields of precision parts inspection and machine-vision measurement.

Resveratrol attenuates autophagy and inflammation after traumatic brain injury by activation of PI3K/Akt/mTOR pathway in rats.

AIM OF THE STUDY: Accumulating studies have demonstrated that neuronal autophagy and inflammation are crucial for hippocampus development in rats subjected to traumatic brain injury (TBI). Therefore, we have investigated whether resveratrol is protective against brain damage through the attenuation of neuronal autophagy and inflammation, and explored underlying mechanisms. MATERIAL AND METHODS: Rats were injected with resveratrol (50 mg/kg, i.p.), following controlled cortical impact (CCI) injury. Brain water content, behavioral studies, and mNSS score were measured to assess the effects of resveratrol treatment. Autophagy-related proteins and inflammatory cytokines in the hippocampus were detected by Western blotting at 12, 24, and 48 hours after TBI. In addition, spatial distribution of LC3 was evaluated with immunofluorescence analysis 24 hours after injury. Finally, factors related to PI3K/Akt/mTOR signaling pathway were assessed at the same time in the hippocampus. RESULTS: Our results depicted that resveratrol could reduce the cerebral edema caused by TBI and improve the recovery of functional deficits in rats. Resveratrol was also able to remarkably reduce the expression of LC3 II and Beclin-1, while increased the expression levels of P62 in the hippocampus. Moreover, we found that interleukin b (IL-1b) and tumor necrosis factor a (TNF-a) were significantly decreased in resveratrol-treated rats. Indeed, we observed an activation of the PI3K/Akt/mTOR pathway after TBI, which may be related to the neuro-protective effect of resveratrol. CONCLUSIONS: Data presented herein support that resveratrol is a potential treatment against TBI through the inhibition of neuronal autophagy and inflammation by activation of PI3K/Akt/mTOR pathway.